107 lines
3.0 KiB
JavaScript
107 lines
3.0 KiB
JavaScript
|
/*
|
||
|
|
MIT License http://www.opensource.org/licenses/mit-license.php
|
||
|
|
Author Tobias Koppers @sokra
|
||
|
|
*/
|
||
|
|
|
||
|
|
"use strict";
|
||
|
|
|
||
|
|
/** @typedef {import("../util/Hash")} Hash */
|
||
|
|
|
||
|
|
/**
|
||
|
|
* StringXor class provides methods for performing
|
||
|
|
* [XOR operations](https://en.wikipedia.org/wiki/Exclusive_or) on strings. In this context
|
||
|
|
* we operating on the character codes of two strings, which are represented as
|
||
|
|
* [Buffer](https://nodejs.org/api/buffer.html) objects.
|
||
|
|
*
|
||
|
|
* We use [StringXor in webpack](https://github.com/webpack/webpack/commit/41a8e2ea483a544c4ccd3e6217bdfb80daffca39)
|
||
|
|
* to create a hash of the current state of the compilation. By XOR'ing the Module hashes, it
|
||
|
|
* doesn't matter if the Module hashes are sorted or not. This is useful because it allows us to avoid sorting the
|
||
|
|
* Module hashes.
|
||
|
|
*
|
||
|
|
* @example
|
||
|
|
* ```js
|
||
|
|
* const xor = new StringXor();
|
||
|
|
* xor.add('hello');
|
||
|
|
* xor.add('world');
|
||
|
|
* console.log(xor.toString());
|
||
|
|
* ```
|
||
|
|
*
|
||
|
|
* @example
|
||
|
|
* ```js
|
||
|
|
* const xor = new StringXor();
|
||
|
|
* xor.add('foo');
|
||
|
|
* xor.add('bar');
|
||
|
|
* const hash = createHash('sha256');
|
||
|
|
* hash.update(xor.toString());
|
||
|
|
* console.log(hash.digest('hex'));
|
||
|
|
* ```
|
||
|
|
*/
|
||
|
|
class StringXor {
|
||
|
|
constructor() {
|
||
|
|
/** @type {Buffer|undefined} */
|
||
|
|
this._value = undefined;
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Adds a string to the current StringXor object.
|
||
|
|
*
|
||
|
|
* @param {string} str string
|
||
|
|
* @returns {void}
|
||
|
|
*/
|
||
|
|
add(str) {
|
||
|
|
const len = str.length;
|
||
|
|
const value = this._value;
|
||
|
|
if (value === undefined) {
|
||
|
|
/**
|
||
|
|
* We are choosing to use Buffer.allocUnsafe() because it is often faster than Buffer.alloc() because
|
||
|
|
* it allocates a new buffer of the specified size without initializing the memory.
|
||
|
|
*/
|
||
|
|
const newValue = (this._value = Buffer.allocUnsafe(len));
|
||
|
|
for (let i = 0; i < len; i++) {
|
||
|
|
newValue[i] = str.charCodeAt(i);
|
||
|
|
}
|
||
|
|
return;
|
||
|
|
}
|
||
|
|
const valueLen = value.length;
|
||
|
|
if (valueLen < len) {
|
||
|
|
const newValue = (this._value = Buffer.allocUnsafe(len));
|
||
|
|
let i;
|
||
|
|
for (i = 0; i < valueLen; i++) {
|
||
|
|
newValue[i] = value[i] ^ str.charCodeAt(i);
|
||
|
|
}
|
||
|
|
for (; i < len; i++) {
|
||
|
|
newValue[i] = str.charCodeAt(i);
|
||
|
|
}
|
||
|
|
} else {
|
||
|
|
for (let i = 0; i < len; i++) {
|
||
|
|
value[i] = value[i] ^ str.charCodeAt(i);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns a string that represents the current state of the StringXor object. We chose to use "latin1" encoding
|
||
|
|
* here because "latin1" encoding is a single-byte encoding that can represent all characters in the
|
||
|
|
* [ISO-8859-1 character set](https://en.wikipedia.org/wiki/ISO/IEC_8859-1). This is useful when working
|
||
|
|
* with binary data that needs to be represented as a string.
|
||
|
|
*
|
||
|
|
* @returns {string} Returns a string that represents the current state of the StringXor object.
|
||
|
|
*/
|
||
|
|
toString() {
|
||
|
|
const value = this._value;
|
||
|
|
return value === undefined ? "" : value.toString("latin1");
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Updates the hash with the current state of the StringXor object.
|
||
|
|
*
|
||
|
|
* @param {Hash} hash Hash instance
|
||
|
|
*/
|
||
|
|
updateHash(hash) {
|
||
|
|
const value = this._value;
|
||
|
|
if (value !== undefined) hash.update(value);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
module.exports = StringXor;
|